usr/src/cmd/sed/main.c

   1 /*
   2  * Copyright 2010 Nexenta Systems, Inc.  All rights reserved.
   3  * Copyright (c) 1992 Diomidis Spinellis.
   4  * Copyright (c) 1992, 1993
   5  *      The Regents of the University of California.  All rights reserved.
   6  *
   7  * This code is derived from software contributed to Berkeley by
   8  * Diomidis Spinellis of Imperial College, University of London.
   9  *
  10  * Redistribution and use in source and binary forms, with or without
  11  * modification, are permitted provided that the following conditions
  12  * are met:
  13  * 1. Redistributions of source code must retain the above copyright
  14  *    notice, this list of conditions and the following disclaimer.
  15  * 2. Redistributions in binary form must reproduce the above copyright
  16  *    notice, this list of conditions and the following disclaimer in the
  17  *    documentation and/or other materials provided with the distribution.
  18  * 4. Neither the name of the University nor the names of its contributors
  19  *    may be used to endorse or promote products derived from this software
  20  *    without specific prior written permission.
  21  *
  22  * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
  23  * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
  24  * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
  25  * ARE DISCLAIMED.  IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
  26  * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
  27  * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
  28  * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
  29  * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
  30  * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
  31  * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
  32  * SUCH DAMAGE.
  33  */
  34
  35 #include <sys/types.h>
  36 #include <sys/mman.h>
  37 #include <sys/param.h>
  38 #include <sys/stat.h>
  39
  40 #include <err.h>
  41 #include <errno.h>
  42 #include <fcntl.h>
  43 #include <libgen.h>
  44 #include <limits.h>
  45 #include <locale.h>
  46 #include <regex.h>
  47 #include <stddef.h>
  48 #include <stdio.h>
  49 #include <stdlib.h>
  50 #include <string.h>
  51 #include <unistd.h>
  52 #include <libintl.h>
  53
  54 #include "defs.h"
  55 #include "extern.h"
  56
  57 /*
  58  * Linked list of units (strings and files) to be compiled
  59  */
  60 struct s_compunit {
  61         struct s_compunit *next;
  62         enum e_cut {CU_FILE, CU_STRING} type;
  63         char *s;                        /* Pointer to string or fname */
  64 };
  65
  66 /*
  67  * Linked list pointer to compilation units and pointer to current
  68  * next pointer.
  69  */
  70 static struct s_compunit *script, **cu_nextp = &script;
  71
  72 /*
  73  * Linked list of files to be processed
  74  */
  75 struct s_flist {
  76         char *fname;
  77         struct s_flist *next;
  78 };
  79
  80 /*
  81  * Linked list pointer to files and pointer to current
  82  * next pointer.
  83  */
  84 static struct s_flist *files, **fl_nextp = &files;
  85
  86 FILE *infile;                   /* Current input file */
  87 FILE *outfile;                  /* Current output file */
  88
  89 int aflag, eflag, nflag;
  90 int rflags = 0;
  91 static int rval;                /* Exit status */
  92
  93 static int ispan;               /* Whether inplace editing spans across files */
  94
  95 /*
  96  * Current file and line number; line numbers restart across compilation
  97  * units, but span across input files.  The latter is optional if editing
  98  * in place.
  99  */
 100 const char *fname;              /* File name. */
 101 const char *outfname;           /* Output file name */
 102 static char oldfname[PATH_MAX]; /* Old file name (for in-place editing) */
 103 static char tmpfname[PATH_MAX]; /* Temporary file name (for in-place editing) */
 104 static const char *inplace;     /* Inplace edit file extension. */
 105 ulong_t linenum;
 106
 107 static void add_compunit(enum e_cut, char *);
 108 static void add_file(char *);
 109 static void usage(void);
 110 static char *getln(FILE *, size_t *);
 111
 112
 113 int
 114 main(int argc, char *argv[])
 115 {
 116         int c, fflag;
 117         char *temp_arg;
 118
 119         (void) setlocale(LC_ALL, "");
 120
 121 #ifndef TEXT_DOMAIN
 122 #define TEXT_DOMAIN     "SYS_TEST"
 123 #endif
 124         (void) textdomain(TEXT_DOMAIN);
 125
 126         fflag = 0;
 127         inplace = NULL;
 128
 129         while ((c = getopt(argc, argv, "EI:ae:f:i:lnr")) != -1)
 130                 switch (c) {
 131                 case 'r':               /* Gnu sed compat */
 132                 case 'E':
 133                         rflags = REG_EXTENDED;
 134                         break;
 135                 case 'I':
 136                         inplace = optarg;
 137                         ispan = 1;      /* span across input files */
 138                         break;
 139                 case 'a':
 140                         aflag = 1;
 141                         break;
 142                 case 'e':
 143                         eflag = 1;
 144                         if (asprintf(&temp_arg, "%s\n", optarg) <= 1)
 145                                 err(1, "asprintf");
 146                         add_compunit(CU_STRING, temp_arg);
 147                         break;
 148                 case 'f':
 149                         fflag = 1;
 150                         add_compunit(CU_FILE, optarg);
 151                         break;
 152                 case 'i':
 153                         inplace = optarg;
 154                         ispan = 0;      /* don't span across input files */
 155                         break;
 156                 case 'l':
 157                         /* On SunOS, setlinebuf "returns no useful value */
 158                         (void) setlinebuf(stdout);
 159                         break;
 160                 case 'n':
 161                         nflag = 1;
 162                         break;
 163                 default:
 164                 case '?':
 165                         usage();
 166                 }
 167         argc -= optind;
 168         argv += optind;
 169
 170         /* First usage case; script is the first arg */
 171         if (!eflag && !fflag && *argv) {
 172                 add_compunit(CU_STRING, *argv);
 173                 argv++;
 174         }
 175
 176         compile();
 177
 178         /* Continue with first and start second usage */
 179         if (*argv)
 180                 for (; *argv; argv++)
 181                         add_file(*argv);
 182         else
 183                 add_file(NULL);
 184         process();
 185         cfclose(prog, NULL);
 186         if (fclose(stdout))
 187                 err(1, "stdout");
 188         return (rval);
 189 }
 190
 191 static void
 192 usage(void)
 193 {
 194         (void) fputs(_("usage: sed script [-Ealn] [-i extension] [file ...]\n"
 195             "       sed [-Ealn] [-i extension] [-e script] ... "
 196             "[-f script_file] ... [file ...]"),
 197             stderr);
 198         exit(1);
 199 }
 200
 201 /*
 202  * Like fgets, but go through the chain of compilation units chaining them
 203  * together.  Empty strings and files are ignored.
 204  */
 205 char *
 206 cu_fgets(char *buf, int n, int *more)
 207 {
 208         static enum {ST_EOF, ST_FILE, ST_STRING} state = ST_EOF;
 209         static FILE *f;         /* Current open file */
 210         static char *s;         /* Current pointer inside string */
 211         static char string_ident[30];
 212         char *p;
 213
 214 again:
 215         switch (state) {
 216         case ST_EOF:
 217                 if (script == NULL) {
 218                         if (more != NULL)
 219                                 *more = 0;
 220                         return (NULL);
 221                 }
 222                 linenum = 0;
 223                 switch (script->type) {
 224                 case CU_FILE:
 225                         if ((f = fopen(script->s, "r")) == NULL)
 226                                 err(1, "%s", script->s);
 227                         fname = script->s;
 228                         state = ST_FILE;
 229                         goto again;
 230                 case CU_STRING:
 231                         if (((size_t)snprintf(string_ident,
 232                             sizeof (string_ident), "\"%s\"", script->s)) >=
 233                             sizeof (string_ident) - 1)
 234                                 (void) strcpy(string_ident +
 235                                     sizeof (string_ident) - 6, " ...\"");
 236                         fname = string_ident;
 237                         s = script->s;
 238                         state = ST_STRING;
 239                         goto again;
 240                 }
 241                 /*NOTREACHED*/
 242
 243         case ST_FILE:
 244                 if ((p = fgets(buf, n, f)) != NULL) {
 245                         linenum++;
 246                         if (linenum == 1 && buf[0] == '#' && buf[1] == 'n')
 247                                 nflag = 1;
 248                         if (more != NULL)
 249                                 *more = !feof(f);
 250                         return (p);
 251                 }
 252                 script = script->next;
 253                 (void) fclose(f);
 254                 state = ST_EOF;
 255                 goto again;
 256         case ST_STRING:
 257                 if (linenum == 0 && s[0] == '#' && s[1] == 'n')
 258                         nflag = 1;
 259                 p = buf;
 260                 for (;;) {
 261                         if (n-- <= 1) {
 262                                 *p = '\0';
 263                                 linenum++;
 264                                 if (more != NULL)
 265                                         *more = 1;
 266                                 return (buf);
 267                         }
 268                         switch (*s) {
 269                         case '\0':
 270                                 state = ST_EOF;
 271                                 if (s == script->s) {
 272                                         script = script->next;
 273                                         goto again;
 274                                 } else {
 275                                         script = script->next;
 276                                         *p = '\0';
 277                                         linenum++;
 278                                         if (more != NULL)
 279                                                 *more = 0;
 280                                         return (buf);
 281                                 }
 282                         case '\n':
 283                                 *p++ = '\n';
 284                                 *p = '\0';
 285                                 s++;
 286                                 linenum++;
 287                                 if (more != NULL)
 288                                         *more = 0;
 289                                 return (buf);
 290                         default:
 291                                 *p++ = *s++;
 292                         }
 293                 }
 294         }
 295         /* NOTREACHED */
 296         return (NULL);
 297 }
 298
 299 /*
 300  * Like fgets, but go through the list of files chaining them together.
 301  * Set len to the length of the line.
 302  */
 303 int
 304 mf_fgets(SPACE *sp, enum e_spflag spflag)
 305 {
 306         struct stat sb;
 307         size_t len;
 308         char *p;
 309         int c;
 310         static int firstfile;
 311
 312         if (infile == NULL) {
 313                 /* stdin? */
 314                 if (files->fname == NULL) {
 315                         if (inplace != NULL)
 316                                 errx(1,
 317                                     _("-I or -i may not be used with stdin"));
 318                         infile = stdin;
 319                         fname = "stdin";
 320                         outfile = stdout;
 321                         outfname = "stdout";
 322                 }
 323                 firstfile = 1;
 324         }
 325
 326         for (;;) {
 327                 if (infile != NULL && (c = getc(infile)) != EOF) {
 328                         (void) ungetc(c, infile);
 329                         break;
 330                 }
 331                 /* If we are here then either eof or no files are open yet */
 332                 if (infile == stdin) {
 333                         sp->len = 0;
 334                         return (0);
 335                 }
 336                 if (infile != NULL) {
 337                         (void) fclose(infile);
 338                         if (*oldfname != '\0') {
 339                                 /* if there was a backup file, remove it */
 340                                 (void) unlink(oldfname);
 341                                 if (link(fname, oldfname) != 0) {
 342                                         warn("link()");
 343                                         (void) unlink(tmpfname);
 344                                         exit(1);
 345                                 }
 346                                 *oldfname = '\0';
 347                         }
 348                         if (*tmpfname != '\0') {
 349                                 if (outfile != NULL && outfile != stdout)
 350                                         if (fclose(outfile) != 0) {
 351                                                 warn("fclose()");
 352                                                 (void) unlink(tmpfname);
 353                                                 exit(1);
 354                                         }
 355                                 outfile = NULL;
 356                                 if (rename(tmpfname, fname) != 0) {
 357                                         /* this should not happen really! */
 358                                         warn("rename()");
 359                                         (void) unlink(tmpfname);
 360                                         exit(1);
 361                                 }
 362                                 *tmpfname = '\0';
 363                         }
 364                         outfname = NULL;
 365                 }
 366                 if (firstfile == 0)
 367                         files = files->next;
 368                 else
 369                         firstfile = 0;
 370                 if (files == NULL) {
 371                         sp->len = 0;
 372                         return (0);
 373                 }
 374                 fname = files->fname;
 375                 if (inplace != NULL) {
 376                         char bn[PATH_MAX];
 377                         char dn[PATH_MAX];
 378                         (void) strlcpy(bn, fname, sizeof (bn));
 379                         (void) strlcpy(dn, fname, sizeof (dn));
 380                         if (lstat(fname, &sb) != 0)
 381                                 err(1, "%s", fname);
 382                         if (!(sb.st_mode & S_IFREG))
 383                                 fatal(_("in-place editing only "
 384                                     "works for regular files"));
 385                         if (*inplace != '\0') {
 386                                 (void) strlcpy(oldfname, fname,
 387                                     sizeof (oldfname));
 388                                 len = strlcat(oldfname, inplace,
 389                                     sizeof (oldfname));
 390                                 if (len > sizeof (oldfname))
 391                                         fatal(_("name too long"));
 392                         }
 393                         len = snprintf(tmpfname, sizeof (tmpfname),
 394                             "%s/.!%ld!%s", dirname(dn), (long)getpid(),
 395                             basename(bn));
 396                         if (len >= sizeof (tmpfname))
 397                                 fatal(_("name too long"));
 398                         (void) unlink(tmpfname);
 399                         if ((outfile = fopen(tmpfname, "w")) == NULL)
 400                                 err(1, "%s", fname);
 401                         if (fchown(fileno(outfile), sb.st_uid, sb.st_gid) != 0)
 402                                 warn("fchown()");
 403                         if (fchmod(fileno(outfile), sb.st_mode & 07777) != 0)
 404                                 warn("fchmod()");
 405                         outfname = tmpfname;
 406                         if (!ispan) {
 407                                 linenum = 0;
 408                                 resetstate();
 409                         }
 410                 } else {
 411                         outfile = stdout;
 412                         outfname = "stdout";
 413                 }
 414                 if ((infile = fopen(fname, "r")) == NULL) {
 415                         warn("%s", fname);
 416                         rval = 1;
 417                         continue;
 418                 }
 419         }
 420         /*
 421          * We are here only when infile is open and we still have something
 422          * to read from it.
 423          *
 424          * Use fgetln so that we can handle essentially infinite input data.
 425          * Can't use the pointer into the stdio buffer as the process space
 426          * because the ungetc() can cause it to move.
 427          */
 428         p = getln(infile, &len);
 429         if (ferror(infile))
 430                 errx(1, "%s: %s", fname, strerror(errno ? errno : EIO));
 431         if (len != 0 && p[len - 1] == '\n')
 432                 len--;
 433         cspace(sp, p, len, spflag);
 434
 435         linenum++;
 436
 437         return (1);
 438 }
 439
 440 /*
 441  * Add a compilation unit to the linked list
 442  */
 443 static void
 444 add_compunit(enum e_cut type, char *s)
 445 {
 446         struct s_compunit *cu;
 447
 448         if ((cu = malloc(sizeof (struct s_compunit))) == NULL)
 449                 err(1, "malloc");
 450         cu->type = type;
 451         cu->s = s;
 452         cu->next = NULL;
 453         *cu_nextp = cu;
 454         cu_nextp = &cu->next;
 455 }
 456
 457 /*
 458  * Add a file to the linked list
 459  */
 460 static void
 461 add_file(char *s)
 462 {
 463         struct s_flist *fp;
 464
 465         if ((fp = malloc(sizeof (struct s_flist))) == NULL)
 466                 err(1, "malloc");
 467         fp->next = NULL;
 468         *fl_nextp = fp;
 469         fp->fname = s;
 470         fl_nextp = &fp->next;
 471 }
 472
 473 int
 474 lastline(void)
 475 {
 476         int ch;
 477
 478         if (files->next != NULL && (inplace == NULL || ispan))
 479                 return (0);
 480         if ((ch = getc(infile)) == EOF)
 481                 return (1);
 482         (void) ungetc(ch, infile);
 483         return (0);
 484 }
 485
 486 char *
 487 getln(FILE *in, size_t *lenp)
 488 {
 489         static char     *buffer = NULL;
 490         static size_t   sz = 0;
 491
 492         size_t          len = 0;
 493
 494         for (;;) {
 495                 if (sz <= (len + 1)) {
 496                         char *nb;
 497                         if ((nb = realloc(buffer, sz + LINE_MAX)) == NULL) {
 498                                 err(1, "realloc");
 499                         }
 500                         buffer = nb;
 501                         sz += LINE_MAX;
 502                 }
 503
 504                 buffer[len] = 0;
 505
 506                 if (fgets(buffer + len, sz - len, in) == NULL) {
 507                         /* END OF FILE */
 508                         *lenp = len;
 509                         break;
 510                 }
 511
 512                 len += strlen(buffer + len);
 513
 514                 if (buffer[len - 1] == '\n') {
 515                         /* got the new line */
 516                         *lenp = len;
 517                         break;
 518                 }
 519         }
 520
 521         return (buffer);
 522 }